for those that use AIM loggers
 

have you ever seen a RPM glitch like that? I had that in a couple of laps and never had that before with any other car.


second question, what kind of oil pressure do you guys get, the lowest value i mean? While running i had pressure as low as 20 and in the pits it drops under 10. Is that normal? I never went under 60 with my Honda.


http://img.photobucket.com/albums/v3...psf55007a7.jpg

it's over 9000!!




no, I've never had something like that :o

No idea about oil pressure since I don't have a sensor installed yet.

As for the weird RPM readings, yeah, I've seen them too (and no, I didn't money shift; I had one while going five miles per hour in pits). If it was a single data point I would think it was just a bad message, but these are several samples wide with a definite rise and fall. It is possible that there is some interpolation going on though which could explain that, but the thought of corrupted CAN bus messages, even on the diagnostic bus, is frightening.

Another possibility (which is what I think is highly probable) is that AIM's interpretation of the CAN message is incomplete. It is possible that they (Subaru/Toyota) are using a high order bit to signal some variant condition on the RPM signal while AIM is interpreting it as simply a higher RPM count. If the message were composed of 15 bits, then bit 15 being high would add 16384 to the value of the normal RPM bits. In this case, if your RPM is actually 6727 (as your other line with about the same shape suggests), then 23525 - 6727 = 16798 which is damn close to 16384. This seems to make sense for me.
I wouldn't be surprised if this meant something like knock detection or some such.

Is the pressure still rising/falling with RPM?


The pressure will naturally drop at the end of the session since the oil is HOT and you're now in low RPM coasting back to your parking spot.


My old Miata used to see below 10 after a hard session on track idling back through the pits.

Is your car an early 2012 build? The early models have a... questionable OBD2 port that makes the connection shakey.

Mine is a MY2014 and I am still seeing this behavior. Also the high bit with other meaning makes an awful lot of sense.

It does... but I've actually never seen that before.

Strange.

it is a MXL so it is hard wired in the car so no loose connection


ddeflyer, i didn't get a word of what you are saying, can you please share how you find the number 16384?


I have no clue how those can communication work by the way so explain it as you would to a complete noob :D


thanks

16384 expressed in binary is 100000000000000, the last 14 digits is all you need to express any realistic RPM values for these cars (2^14-1 = 16383 + "0" expressible in 14 bits). Since we can assume that there are at least 15 bits in the data section of the CAN bus message (otherwise you couldn't have values as high as we are seeing), there are at least one bit that usable (and I suspect being used) for flagging or transmitting additional information. Frankly I am kind of guessing here since I don't know how many bits are in the message's data section.

Another possibility that can't be ignored is that the message is being given as a 16 bit number where all listeners just ignore the first two bits and the message originator is not blanking the target memory when copying a 14 bit value. This would be shitty programming, but not unheard of.